• Rare
• One mutation with large effect
• Discernable mode of inheritance
• Example: Cystic Fibrosis, Sickle Cell Anemia

• Common
• Several polymorphisms with a small effect
• Aggregation
• Exampl: Diabetes Mellitus

• is there evidence of clustering of disease in families?
• how much of the disease occurence in families is due to genetics vs. common environment?
• is there evidence for a major gene? dominant or recessive?
• where are the major genes located?
• where is the exact location of the gene(s), and which polymorphism is associated with the disease?
• what is the true functional variant and how does it cause disease

Dichotomous (usually disease) traits

1. Find out the general population prevalence of the disorder.
2. Count the number of relatives of affected individuals with and without the disease and calculate a prevalence of disease among the relatives
3. Calculate the relative risk ratio λ_r = prevalence in relatives / general population prevalence

OR:

1. Collect cases with the disease
2. Collect well-matched controls without the disease
3. Calculate the ratio of prevalence in the case relatives vs. control relatives

1. Collect monozygotic (MZ) and dizygotic (DZ) twins
2. Ask the question:  Is concordance higher in MZ than DZ twins?

Yes -->evidence for genetic component

1. If MZ concordance is <100% --> nongenetic component involved

• To determine if inheritance is consistent with effect of a major gene
• If major gene exists, what is inheritance pattern?
• Provides supposr but not proof of a genetic effect

• Definition: tendency of alleles close together on the same chormosome to be transmitted together
• Violation of Mendel's Law of Independent Assorment

1. Recruit affected individuals and their adoptive and biological relatives
2. Are biological relatives more likely to be affected than adoptive relatives?

• to look for evidence of a major gene and its inheritance program

1. Informally:  Look for dominant, recessive, or X-linked inheritance in a pedigree
2. Formally
   1. Simple:  Use a statistical test to compare expected proportion of affected offspring from a particular mating type for a particular inheritance pattern (e.g., expect 50% of offspring of Aa x aa mating to be affected with an autosomal dominant disease)
   2. Complex:  Use computer modeling to incorporate effects of multiple genes and environmental effects

• A set of alleles at different loci on the same chromosome

• used after segregation analysis
• test evidence for segregation of a particular marker with disease at particular recombination fractions

• measurement of the distance between loci
• probability of a recombination event between the two loci
• Measured in centimorgans (1 centimorgan = 1% recombination = approximately (varies) 1,000,000 base pairs
• recombinant offspring: parental haplotype disrupted by crossing over (more likely at greater distances)
• non-recombinant offspring: parental haplotype intact

θ = 0

Complete linkage

No observed recombinant offspring

Marker is disease locus or very close

θ = 0.5

No linkage

50% of offspring are recombinants

θ < 0.5

Some linkage

< 50% of offspring are recombinants

Marker may be near locus

This method does not require known inheritance pattern or other parameters including disease prevalence, penetrance or phenocopy rate and does not that segregation analysis be performed beforehand.  However, because fewer assumptions are made, power is lower.  Nonparametric linkage analysis for qualitative (generally disease) looks to see if pairs of relatives who are both affected by a condition are more genetically alike at a given locus than would be expected by chance based on their relationship.  For quantitative traits, the idea is to see if there is a positive correlation between the trait and the amount of alleles shared at a locus.

This method requires a known inheritance pattern and other parameters including disease prevalence, penetrance and phenocopy rate and is thus generally done after segregation analysis and for single gene disorders.

1. Collect families with multiple members having the disease
2. Genotype markers in different regions of the genome
3. Count the number of individuals in which the marker genotype and disease status recombined and did not recombine
4. Calculate the likelihood of having that particular number of recombinants and that particular number of nonrecombinants at a given actual recombination fraction
5. Calculate the likelihood of having that particular number of recombinants and that particular number of nonrecombinants if the true recombination fraction is 0.5 (random, no linkage)
6. Divide the likelihood of linkage by the likelihood of no linkage.  This is the odds of linkage vs. no linkage.
7. Take the log₁₀ of the odds.  This result is the LOD score.  We use the log because (1) it makes the number less unwieldy (e.g., 3 vs. 1000) and (2) it makes it easy to combine LOD scores from different families—just add them together.

1.       Collect pairs of siblings where both members of the sibship have the disease, and if possible, their parents.

2.       Genotype siblings and if possible, parents, for markers located throughout the genome.

3.       At each marker, count/estimate the number of alleles shared IBD for each sib pair.

4.       At each marker, calculate the total proportion of alleles shared IBD.

5.       If the proportion is more than the expected 0.5 for siblings, there is evidence for linkage at that locus.